View Single Post
Old 12-10-2017, 11:14 AM   #21
Katsunami
Grand Sorcerer
Katsunami ought to be getting tired of karma fortunes by now.Katsunami ought to be getting tired of karma fortunes by now.Katsunami ought to be getting tired of karma fortunes by now.Katsunami ought to be getting tired of karma fortunes by now.Katsunami ought to be getting tired of karma fortunes by now.Katsunami ought to be getting tired of karma fortunes by now.Katsunami ought to be getting tired of karma fortunes by now.Katsunami ought to be getting tired of karma fortunes by now.Katsunami ought to be getting tired of karma fortunes by now.Katsunami ought to be getting tired of karma fortunes by now.Katsunami ought to be getting tired of karma fortunes by now.
 
Katsunami's Avatar
 
Posts: 6,111
Karma: 34000001
Join Date: Mar 2008
Device: KPW1, KA1
Quote:
Originally Posted by pdurrant View Post
I was going to complain about your timeline, and then I found, to my surprise that Pascal (1970) did come before C (1972).
Hehe... I know both Pascal and C very well. I've started out on Turbo Pascal 3 (on my XT, in 1990/1991), and then Borland Pascal 7 (on my 486, in 1994).

The original Pascal was extremely structured; it had procedures _and_ functions, and a function without a return value wasn't allowed, typecasting wasn't possible, and it didn't have any pointers, and couldn't use the CPU's registers. Inline assembly wasn't possible. I think you couldn't even do expressions during a return.

I don't know when Turbo/Borland Pascal gained those capabilities, but when I was good enough at programming to use them, version 7 did have all of that. I stuck with it until 1996 or so, writing my own command-line utilities or front-ends for things like PKZIP/UNZIP and ARJ, and then switched to Delphi 4.

I -did- discover C and C++ somewhere around 1995, but I found it to be a weird and cryptic version of Pascal. All of the touted benefits (low level access, assembly, pointer arithmetic and so on) were moot points because Borland Pascal had all of those as well. And Pascal had objects as well if you wanted them. (But they were comparing standard Pascal to standard C of course.)

In the end I switched to C/C++ purely because it had become the standard language (and I did so by using Borland C++ Builder for Windows... HAH!), not because I missed anything in Pascal.

For Windows programming, I finally came into contact with .NET and Visual Studio in 2006 (!) after Borland sold all its stuff and Delphi/Builder were overhauled by other companies, and switching to a new version would basically be the same as switching to a completely different system.

Now, when I still use C or C++, it's in embedded software engineering, mostly on realtime operating systems, or small things on Linux. I don't write applications in it. I've been looking into things like Electron/NodeJS and other web technologies for that, that take HTML/CSS and Javascript out of the browser. Most things that I make and have a user interface don't require the speed of C/C++.

Same with Mercurial by the way. I started out with that in 2006, because Git, as could be expected from it, being written by Linus Torvalds, didn't run on Windows. It started to change in 2012, but I stuck with Mercurial. I recently switched to Git only because it has about 75% market share, and many companies put it down as a requirement (and not Mercurial), so I studied Git and put it onto my CV.

And what's the verdict? As I expected... for day to day operations, it's EXACTLY THE SAME as Mercurial. Maybe it can do some weird stuff Mercurial can't, but if so, I haven't needed it yet. Same as it was with Borland Pascal and C/C++... same stuff, different syntax.
Katsunami is offline   Reply With Quote